SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Strategies & Market Trends : Gorilla and King Portfolio Candidates -- Ignore unavailable to you. Want to Upgrade?


To: BDR who wrote (37040)12/26/2000 1:07:29 PM
From: tinkershaw  Read Replies (1) | Respond to of 54805
 
So each destination (server?) will be assigned a unique frequency? How many servers are there on the net? More than the thousand discrete frequencies into which Avanex claims to be able to divide light. How do they deal with that?

Dale,

Can't say my Duke MBA is helping here either, but Mr. Cao states that 250,000 lambdas would be sufficient to support a network of over 20 million users.

Yes, only a 1000 wavelengths on a strand are currently possible, but the working assumption (as it is for microchips is that Caos law of doubling fiber capacity every 6-9 months will function as consistently as Moore's law of doubling every 18 months - which by the way Intel seems to be exceeding these days with P IV and Rambus :) ).

So at this pace of development 100,000+ channels on fiber is about 3-4 years away. Whether or not Caos law will prove as workable as Moore's law remains to be seen.

Tinker



To: BDR who wrote (37040)1/3/2001 2:37:41 PM
From: BDR  Respond to of 54805
 
Re: MEMS

Tinkershaw and I were discussing switching and MEMS last week. There is an article in RedHerring that describes MEMS and their uses in optical switching. It is pretty basic (even I could understand it) so it may not be useful to everyone but here is the link:

MEMS are microscopic marvels
redherring.com